lnbest0707-uber commented on PR #15234:
URL: https://github.com/apache/pinot/pull/15234#issuecomment-2735020908

   > This won't work for pauseless consumption because segment build happens 
after commit. cc @KKcorps @9aman
   > 
   > Ideally we should try to follow the mechanism introduced in #14479 to 
early commit the segment when it cannot consume more rows
   
   I have a follow up PR for early commit in 
https://github.com/apache/pinot/pull/15120. However, the threshold is not easy 
to set. Concerns are listed in that PR.
   This retry mechanism (enabled by the table config) could be preferred in 
some cases. E.g. for large tables, false alarm on early commits is not always 
acceptable(creating too many segments).
   
   Regarding to pauseless consumption, this issue could also arise and be even 
more severe(no way to recover). In my opinion, we should introduce similar 
thing in the `/reingest` failure recovery loop introduced in 
https://github.com/apache/pinot/pull/14920.


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org
For additional commands, e-mail: commits-h...@pinot.apache.org

Reply via email to